Date:   Sat, 13 Nov 2021 14:01:19 +0800
From:   Herbert Xu <herbert@gondor.apana.org.au>
To:     Kai Ye <yekai13@huawei.com>
Cc:     linux-crypto@vger.kernel.org, linux-kernel@vger.kernel.org,
        wangzhou1@hisilicon.com, liulongfang@huawei.com
Subject: Re: [PATCH v3 2/6] crypto: hisilicon/sec - add ahash alg features
 for Kunpeng920
Message-ID: <20211113060119.GA21020@gondor.apana.org.au>
References: <20211022091055.15369-1-yekai13@huawei.com>
 <20211022091055.15369-3-yekai13@huawei.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20211022091055.15369-3-yekai13@huawei.com>
User-Agent: Mutt/1.10.1 (2018-07-13)
Precedence: bulk

On Fri, Oct 22, 2021 at 05:10:51PM +0800, Kai Ye wrote:
>
> +static int sec_ahash_export(struct ahash_request *req, void *out)
> +{
> +	/*
> +	 * This function dumps the
> +	 * entire state of the ongoing transformation into a provided block of
> +	 * data so it can be @import'ed back later on.
> +	 */
> +	struct sec_req *sreq = ahash_request_ctx(req);
> +	struct sec_auth_ctx *a_ctx = &sreq->ctx->a_ctx;
> +	u8 mac_len = a_ctx->mac_len;
> +
> +	memcpy(out, &sreq->hash_req, sizeof(struct sec_ahash_req));

So this structure seems to contain a bunch of metadata but not
the actual hash state.

> +	/* Export the partial hash data */
> +	memcpy(out + sizeof(struct sec_ahash_req), a_ctx->metamac, mac_len);

This looks like it could be the hash state, but mac_len is clamped
at digest size so it's way too small to contain the entire hash
state.

How is this supposed to work?

Thanks,
-- 
Email: Herbert Xu <herbert@gondor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt